AI Glossary
AI Glossary
Explore our AI Glossary to gain insights into artificial intelligence, machine learning, computer vision, and more. Discover key terms and definitions to navigate the world of AI technologies.
Artificial General Intelligence (AGI)
A theoretical form of AI that possesses the ability to understand, learn, and apply knowledge across a wide variety of tasks, indistinguishable from the human mind.
AI Agent
An autonomous system that can perceive its environment, reason about it, and take actions to achieve specific goals with limited human intervention.
Algorithm
A set of rules or instructions given to an AI, neural network, or computer program to help it learn on its own and solve problems.
AI Alignment
The process of ensuring that artificial intelligence systems act in ways that are consistent with human values, goals, and safety standards.
AI Bias
Systematic errors in artificial intelligence systems that result in unfair or prejudiced outcomes, often stemming from skewed or incomplete training data during the development process.
Backpropagation
The primary algorithm used to train neural networks by calculating errors and adjusting the model’s internal weights to improve accuracy.
Big Data
Extremely large datasets that may be analyzed computationally to reveal patterns, trends, and associations, especially relating to human behavior.
Computer Vision
A field of AI that enables computers to interpret and understand visual information from the world, such as images and videos.
Chatbot
A software application used to conduct an online chat conversation via text or text-to-speech, often powered by LLMs.
Cognitive Computing
AI systems that simulate human thought processes to help solve complex problems in uncertain and ambiguous environments.
Deep Learning
A subset of machine learning based on artificial neural networks with many layers, used for complex tasks like speech and image recognition.
Data Science
An interdisciplinary field that uses scientific methods, processes, algorithms, and systems to extract knowledge and insights from data.
Ethical AI
Guidelines and social norms intended to ensure that AI technologies are developed and used responsibly and fairly.
Embeddings
A technique in NLP where words or phrases are converted into vectors of real numbers, allowing machines to understand semantic similarity.
Fine-Tuning
The process of taking a pre-trained AI model and training it further on a specific dataset to adapt it for a more specialized task.
Generative AI
AI systems capable of creating new content, such as text, images, or audio, based on the patterns they learned from training data.
Hallucination
A confident response by an AI that does not seem to be justified by its training data, resulting in false or nonsensical information.
Inference
The stage where a trained AI model is used to make predictions or solve problems using new, real-world data.
JAX
A specialized library developed by Google for high-performance machine learning research, often used for advanced neural network training and transformations.
Jupyter Notebook
An open-source web application that allows AI developers to create and share documents containing live code, equations, and visualizations.
Keras
An open-source software library that provides a Python interface for artificial neural networks, acting as a user-friendly wrapper for TensorFlow.
Knowledge Graph
A structured representation of information that shows relationships between different entities, used by AI to provide context and improve search results.
K-Nearest Neighbors (KNN)
A simple, versatile machine learning algorithm used for classification and regression by identifying the most similar data points in a dataset.
K-Fold Cross-Validation
A data-splitting technique used to evaluate the performance of an AI model by dividing the data into “K” parts to ensure the results are consistent and reliable.
Large Language Model (LLM)
An AI model trained on massive text datasets, capable of generating, summarizing, and translating human-like language.
Latency
The time it takes for an AI system to process an input and generate a response, critical for real-time applications like voice assistants.
Machine Learning (ML)
A field of AI focused on building systems that learn from data and improve their performance without being explicitly programmed.
Multimodal AI
An AI system capable of processing and understanding multiple types of data simultaneously, such as text, images, and audio.
Natural Language Processing (NLP)
A branch of AI that helps computers understand, interpret, and generate human language in a meaningful way.
Neural Network
A computational model inspired by the human brain’s structure, used to recognize patterns and solve complex problems.
Overfitting
A modeling error that occurs when an AI model learns the training data too well, failing to generalize to new, unseen data.
Prompt Engineering
The art and science of crafting precise inputs to guide generative AI models toward producing high-quality and relevant outputs.
Parameters
The internal variables that an AI model adjusts during training to learn patterns and make accurate predictions.
Q-Learning
A fundamental reinforcement learning algorithm where an AI agent learns the value of an action in a particular state to find the best possible strategy through trial and error.
Quantization
A process used to reduce the size of AI models by converting high-precision numbers into lower-precision ones, allowing complex LLMs to run on consumer hardware with less memory.
Quantum AI
The intersection of quantum computing and artificial intelligence, aiming to use quantum bits (qubits) to solve complex problems much faster than traditional computers.
RAG (Retrieval-Augmented Generation)
A framework that allows LLMs to retrieve facts from an external, reliable knowledge base before generating an answer.
Reinforcement Learning (RL)
A type of machine learning where an agent learns to make decisions by receiving rewards or penalties for its actions.
Supervised Learning
A type of ML where the model is trained on a labeled dataset, meaning the answer is already provided for each example.
Sentiment Analysis
The use of NLP to identify and extract subjective information, such as emotions or opinions, from text data.
Transformer
A neural network architecture that uses self-attention mechanisms, forming the foundation for modern models like GPT and BERT.
Turing Test
A test of a machine’s ability to exhibit intelligent behavior equivalent to, or indistinguishable from, that of a human.
Unsupervised Learning
A type of ML where the model finds hidden patterns or structures in input data without any pre-existing labels.
Uncertainty (AI)
The measure of a model’s lack of confidence in its predictions, often used to determine when a human expert should intervene or when more data is required.
Virtual Assistant
An AI application that uses natural language processing and voice recognition to perform tasks, answer questions, and assist users (e.g., Siri, Alexa).
Vector Database
A specialized type of database that stores data as high-dimensional vectors, enabling fast and efficient similarity searches for RAG and AI applications.
Validation Set
A portion of data held back during AI training to tune model hyperparameters and prevent overfitting before the final testing phase.
Weights (Neural Networks)
Numerical values that represent the strength of connections between neurons in an AI model, which are adjusted during training to minimize errors.
Weak AI
Also known as Narrow AI, this refers to artificial intelligence that is trained and focused on a single, specific task (e.g., facial recognition or translation).
Word2Vec
A popular technique in NLP that uses a neural network to learn word associations from a large body of text, converting words into mathematical vectors.
XAI (Explainable AI)
A set of processes and methods that allows human users to comprehend and trust the results and output created by machine learning algorithms.
Xavier Initialization
A specialized technique for setting the initial weights of a neural network to prevent signals from becoming too small or too large during training.
YOLO (You Only Look Once)
A popular real-time object detection system that can identify multiple objects in an image or video frame with incredible speed and accuracy.
Yield
A term used to describe the efficiency and quality of the output produced by an AI model or a specific dataset during the processing phase.
Zero-Shot Learning
The ability of an AI model to correctly complete a task or recognize a category it was not specifically exposed to during its training phase.
Z-Score (Normalization)
A statistical measurement used in data preprocessing to rescale features so they have a mean of 0 and a standard deviation of 1, helping AI models learn faster.